CN111650878B - Method for optimizing programmability of flow when multiple controllers in software defined network fail - Google Patents

Method for optimizing programmability of flow when multiple controllers in software defined network fail Download PDF

Info

Publication number
CN111650878B
CN111650878B CN202010544094.4A CN202010544094A CN111650878B CN 111650878 B CN111650878 B CN 111650878B CN 202010544094 A CN202010544094 A CN 202010544094A CN 111650878 B CN111650878 B CN 111650878B
Authority
CN
China
Prior art keywords
programmability
flow
offline
path
mapping
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010544094.4A
Other languages
Chinese (zh)
Other versions
CN111650878A (en
Inventor
郭泽华
窦松石
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Institute of Technology BIT
Original Assignee
Beijing Institute of Technology BIT
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Institute of Technology BIT filed Critical Beijing Institute of Technology BIT
Priority to CN202010544094.4A priority Critical patent/CN111650878B/en
Publication of CN111650878A publication Critical patent/CN111650878A/en
Application granted granted Critical
Publication of CN111650878B publication Critical patent/CN111650878B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B19/00—Programme-control systems
    • G05B19/02—Programme-control systems electric
    • G05B19/04—Programme control other than numerical control, i.e. in sequence controllers or logic controllers
    • G05B19/05—Programmable logic controllers, e.g. simulating logic interconnections of signals according to ladder diagrams or function charts
    • G05B19/056—Programming the PLC
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B2219/00—Program-control systems
    • G05B2219/10—Plc systems
    • G05B2219/13—Plc programming
    • G05B2219/13004—Programming the plc

Abstract

The invention discloses a method for optimizing the programmability of flow when multiple controllers in a software defined network fail, which comprises the steps of constructing an optimal flow controller mapping model (OFCM), converting the recovery problem of offline flow in the network into the solving problem of the OFCM, providing a heuristic solution PG for completing the solving of the OFCM, adopting fine-grained flow level remapping to recover the offline flow with lower communication cost when multiple controllers fail, and experiments prove that the method can effectively improve the number of the recovery flow, the programmability of a balance path and the total path programmability of the offline flow under the real topological environment and effectively reduce the communication cost in the recovery process.

Description

Method for optimizing programmability of flow when multiple controllers in software defined network fail
Technical Field
The invention belongs to the technical field of computer networks, and particularly relates to a method for optimizing the programmability of streams when multiple controllers in a software defined network fail.
Background
Maintaining control resiliency is a key issue in applying Software Defined Networking (SDN) to Wide Area Networks (WANs), known as SD-WANs. In SD-WAN, the data plane consists of multiple network domains, each with SDN switches distributed at different physical locations. The control plane has SDN controllers, which are network control software installed in physical servers or virtual machines, for controlling the physical SDN switches within its domain. SDN controllers may fail due to certain unexpected issues (e.g., hardware/software errors, power failures). The failed controller takes all connected switches offline, thereby losing the ability to alter the path of the flows flowing through them, i.e., path programmability, and the flows become offline. Restoring path programmability to offline streams is central to maintaining control flexibility in the event of a controller failure.
In SD-WAN, the core to maintain control plane resiliency is to restore programmability of offline streams in the event of controller failure. This is a complex optimization problem under practical SDN constraints. First, the optimization goal is to maximize the restoration of path programmability, thereby achieving maximum SDN control functionality, and to balance the restored path programmability. Second, the ability of the online controller to restore the offline stream is limited by its processing power. Third, performance metrics (e.g., communication overhead between the switch and the controller) are also believed to provide a fast response to requests from the switch during offline flow restoration.
Existing control resiliency solutions can restore the programmability of offline flows at the switch level. For failed controllers, existing solutions employ a default path programmability recovery solution in OpenFlow to establish a new mapping from offline switches to online controllers. By mapping an offline switch to an online controller, all flows through the switch are controlled by the controller and become programmable.
To efficiently solve this problem, Tanha et al and Killi et al propose solutions to restore the programmability of the paths by mapping the switches to the controllers in a static manner. A static solution is to select and place a standby controller before the controller fails, select the standby controller and map it to the switch. It optimizes network deployment by carefully selecting the location of the controller and the connection between the controller and the switch to reduce the effects of potential controller failures. However, these solutions typically ignore different control loads of the switch and dynamic changes in the control capabilities of the controller. Thus, they are not efficient nor effective in practical environments.
Guo et al propose a dynamic solution to remap offline switches to online controllers in real time by taking into account the status of the switches and controllers at the moment. Although dynamic solutions have been successful in restoring path programmability, two problems remain. First, the path programmability of the recovery streams is unbalanced. Typically only a long and limited number of offline streams can be restored to a programmable state. Secondly, because the current method adopts the coarse-grained level with the switch as the unit for recovery, the effect of recovering the path programmability is not good.
In summary, the method for optimizing the programmability of the offline stream in the prior art mainly has the following problems: first, the recovery granularity is too coarse, usually all flows in the failed switch are recovered; secondly, the dynamic changes of different control loads of the switch and the control capability of the controller are ignored in the recovery process; third, the path programmability of the recovery streams is not balanced.
Disclosure of Invention
In view of this, the present invention provides a method for optimizing the programmability of a flow when multiple controllers fail in a software defined network, which can accurately and efficiently recover the path programmability of an offline flow when multiple controllers fail in the network.
The invention provides a method for optimizing the programmability of streams when multiple controllers in a software defined network fail, which comprises the following steps:
step 1, establishing an optimal flow controller mapping model to describe a mapping relation between an offline flow and an online controller in a network, wherein the optimal flow controller mapping model is shown as the following formula:
Figure BDA0002540067800000031
where r is the minimum value of path programmability for all offline streams; l is the number of the offline stream f, and L is the total number of the offline streams in the network; i is the number of the offline switch, and N is the total number of the offline switches s in the network; j is the serial number of the online controller c, and M is the total number of the online controllers in the network;
Figure BDA0002540067800000032
Dijfor a switch siAnd a controller cjA propagation delay therebetween, λ is a constant greater than or equal to 0,
Figure BDA0002540067800000033
is flowed through siOff-line flow flIn case of code iThe number of paths contained on the machine switch s;
Figure BDA0002540067800000034
is a Boolean type variable when
Figure BDA0002540067800000035
When the value is 1, f is expressedlFlows through siAnd siAt least two paths to flOtherwise
Figure BDA0002540067800000036
The value is 0;
Figure BDA0002540067800000037
is a Boolean type variable when
Figure BDA0002540067800000038
A value of 1 indicates a flow through siF of (a)lMapping to cjWhen is coming into contact with
Figure BDA0002540067800000039
When the value is 0, the flow is indicated to pass through siF of (a)lNot mapped to cj；
Figure BDA00025400678000000310
Is cjThe remaining capacity of (c); max [. X [ ]]Is the maximum value;
step 2, mixing
Figure BDA00025400678000000311
After relaxation of r, solving the optimal flow controller mapping model to obtain a solution set
Figure BDA00025400678000000312
Figure BDA00025400678000000313
Wherein the content of the first and second substances,
Figure BDA00025400678000000314
is formed by flThe k mapping relationships of (a) and (b) form a set, the mapping relationships including mappings between switches and controllers and probabilities of the mappings; selecting the maximum value of k as iteration times T; setting the path programmability of all offline streams to 0; will map the relationship
Figure BDA0002540067800000041
All are set to 0;
step 3, establishing a set L to be tested containing all offline streams*；
Step 4, from the set L to be tested*Selecting an offline flow fl；
Step 5, from flCorresponding to
Figure BDA0002540067800000042
To select the e-th mapping relation
Figure BDA0002540067800000043
Wherein e is more than or equal to 1 and less than or equal to k to obtain
Figure BDA0002540067800000044
The corresponding switch number i and the controller number j, i.e. the mapping relation is
Figure BDA0002540067800000045
Will be provided with
Figure BDA0002540067800000046
The value of (a) is set to 1; will be provided with
Figure BDA0002540067800000047
From
Figure BDA0002540067800000048
Deleting;
if it is
Figure BDA0002540067800000049
Satisfy the requirement of
Figure BDA00025400678000000410
And
Figure BDA00025400678000000411
then
Figure BDA00025400678000000412
To make the mapping feasible, flAccording to
Figure BDA00025400678000000413
Completing the mapping to switches and controllers and updating the total remaining processing power of all controllers in the network, calculating and updating flPath programmability of rlExecuting the step 6;
if it is
Figure BDA00025400678000000414
Not meet the requirements of
Figure BDA00025400678000000415
Or
Figure BDA00025400678000000416
Then will be
Figure BDA00025400678000000417
Setting to 0; if it is
Figure BDA00025400678000000418
If not, changing the value of e and executing the step 5; if it is
Figure BDA00025400678000000419
If it is empty, f islFrom L*Deleting, and executing step 7;
step 6, if the total residual processing capacity is 0, the test is finished and the process is exited; otherwise, will flFrom L*Deleting, and executing step 7;
step 7, if L*If not, executing step 4; otherwise, making T self-reduce by 1, if T is not 0, executing step 3 according to the path programmability of the updated flow; if T is 0, the test is finished and the process is exited.
Further, in said step 5, from flCorresponding to
Figure BDA00025400678000000420
To select the e-th mapping relation
Figure BDA00025400678000000421
And selecting according to a principle that the mapping probability is prior, namely preferentially selecting the mapping relation with higher mapping probability.
Further, in step 7, according to the updated path programmability of the stream, when step 3 is executed in the next iteration, the set L to be tested is first tested*The flows in the sequence are sequenced from large to small according to the programmability of the path; accordingly, the step 4 is from L*The offline stream with the greatest path programmability is selected.
Further, the path programmability is calculated by adopting a programmable path calculation method based on a programmable structure.
Has the advantages that:
the invention converts the recovery problem of the off-line flow in the network into the solving problem of the OFCM model by constructing an optimal flow controller mapping model (OFCM), and provides a heuristic solution PG to complete the solving of the OFCM model, the established model and the solving process thereof, and the invention adopts fine-grained flow level remapping, can recover the off-line flow with lower communication overhead when a plurality of controllers have faults.
Drawings
FIG. 1 is a schematic diagram of the programmability of the flow optimization method for the failure of multiple controllers in the software defined network according to the present invention.
Fig. 2(a) is a schematic diagram 1 of a path programmability calculation process adopted by the method for optimizing the programmability of a flow when multiple controllers fail in a software-defined network according to the present invention.
Fig. 2(b) is a schematic diagram 2 illustrating the path programmability calculation process adopted by the method for optimizing the programmability of the flow when multiple controllers fail in the software-defined network according to the present invention.
Detailed Description
The invention is described in detail below by way of example with reference to the accompanying drawings.
The invention provides a method for optimizing the programmability of streams when multiple controllers fail in a software defined network, which has the core idea that: a method for improving path programmability in a software defined network is used to restore path programmability of offline streams in the network.
The invention provides a method for optimizing the programmability of streams when multiple controllers in a software defined network fail, which specifically comprises the following steps:
step 1, establishing an optimal flow controller mapping model (OFCM) to describe the mapping relation between an offline flow and an online controller in a network, wherein the optimal flow controller mapping model is shown as a formula (1):
Figure BDA0002540067800000061
where r is the minimum value of path programmability for all offline streams; l is the number of the offline stream f, and L is the total number of the offline streams in the network; i is the number of the switch, and N is the total number of the switches s in the network; j is the serial number of the online controller c, and M is the total number of the online controllers in the network;
Figure BDA0002540067800000062
Dijfor a switch siAnd a controller cjA propagation delay therebetween, λ is a constant greater than or equal to 0,
Figure BDA0002540067800000063
is flowed through siOff-line flow flThe number of paths involved;
Figure BDA0002540067800000064
is a Boolean type variable when
Figure BDA0002540067800000065
When the value is 1, f is expressedlFlows through siAnd siAt least two paths to flOtherwise
Figure BDA00025400678000000610
The value is 0;
Figure BDA0002540067800000066
is a Boolean type variable when
Figure BDA0002540067800000067
A value of 1 indicates a flow through siF of (a)lMapping to cjWhen is coming into contact with
Figure BDA0002540067800000068
When the value is 0, the flow is indicated to pass through siF of (a)lNot mapped to cj；
Figure BDA0002540067800000069
Is cjThe remaining capacity of (c).
Typically, a software-defined wide area network (SD-WAN) consists of H controllers located at H locations, each controller controlling a switch domain. Assume that the set of online controllers is C ═ C1,…,Cj,…,CMThe failed controllers are grouped as { C }M+1,…,CHThe set of offline switches is S ═ S1,…,si,…,sNThe failure controller controls the N offline switches. The problem to be solved by the present invention is to map the flows flowing through all offline switches onto the online controller. The set of flows flowing through the offline switch S is F ═ F1,f2,…,fl,…,fL}. If flow flFlow through switch siAnd s andiat least two paths to flIs expressed as
Figure BDA0002540067800000071
Otherwise
Figure BDA0002540067800000072
In the present invention, use is made of
Figure BDA0002540067800000073
To indicate the flow through the switch siFlow f oflMapping to controller Cj(ii) a Otherwise
Figure BDA0002540067800000074
The following describes the constraints of the optimal flow controller mapping model:
1) the flow controller maps the constraints. If the off-line flow flFlow through switch siThen the offline stream can be mapped to at most one online controller, as shown in equation (2):
Figure BDA0002540067800000075
2) controller processing power constraints. When a controller fails to become an offline switch, the online controller controls traffic from the offline switch without interrupting its own normal operation. The control overhead of a controller is equal to the total overhead of controlling its associated streams in its domain, and the processing power of the controller is expressed in terms of the total number of streams it can control without introducing additional delay (e.g., queuing delay). In the present invention, the constraint of the processing capacity of the controller means that the load to be processed of the controller is not greater than the remaining processing capacity thereof, as shown in formula (3):
Figure BDA0002540067800000076
wherein the content of the first and second substances,
Figure BDA0002540067800000077
presentation controller cjThe remaining processing power of.
3) Path programmability constraints for the flow. Statistically, a flow with a long path has higher programmability than a flow with a short path because a long path increases the probability of control of the flow. In other words, the path length virtually prioritizes the streams and causes a path programmability imbalance between the streams. When the flow has a large flow, the routing cannot be restarted to improve the load balancing performance of the network, and the flow size may change continuously with time, so it is a reasonable solution to solve the problem to strive for each offline flow to have the same path programmability. Path programmability of a flow is expressed as the ability of a switch to change the path of a flow flowing through the switch.
Path-programmable computation of flows is a complex problem in restoring any number of switches on the forwarding path of an offline flow. The difficulty is how to eliminate redundant paths between programmable paths of different switches. Aiming at the problem, the invention adopts a programmable path calculation method based on a programmable structure. The programmable fabric is defined as a network fabric formed by recovery switches on an offline flow forwarding path. Based on this structure, a directed network topology can be simplified: finding out the recovered exchanger, its next hop neighbor node and the destination node of the flow; only the shortest path part between the exchangers is replaced by direct connection to form a programmable structure; pruning adjacent nodes, if the shortest path exists between the adjacent nodes and the destination node of the stream, keeping the nodes, otherwise deleting the edges adjacent to the programmable structures of the nodes; the computation programmable structure is considered as a whole, and the number of connections with the outside is the programmability, as shown in fig. 1. For example, the calculation process of path programmability is shown in fig. 2(a), where programmable structure a is s20 and s22, and programmable structure a is connected out three sides: s20-s24, s22-s24 and s22-s25, which are sources of the path programmability of the programmable structure a, that is, the value of the path programmability of the programmable structure a is 3; in fig. 2(B), the programmable structures B are s21 and s22, where s21 is directly connected to s22 instead of the shortest path through s20 in fig. 2(a), and the programmable structures B are externally connected to four sides: s21-s23, s21-s25, s22-s25 and s22-s24, so the path programmability of programmable structure B is 4.
Thus, the path programmability constraint of a flow can be expressed using equation (4):
Figure BDA0002540067800000081
a typical solution to integer programming in the prior art is to use an integer program optimization solver to obtain an optimal solution to the OFCM problem described above. However, as the size of the network increases, the solution space may increase substantially, and finding a viable solution may take a long time or even be impossible. Therefore, in the invention, a heuristic algorithm named PG is provided to solve the problem, thereby improving the performance and time complexity. The key idea of PG is to test and increase the path programmability of each flow by following a certain probability of the flow controller mapping. The specific steps of the PG algorithm include the following calculation process from step 2 to step 7.
Step 2, mixing
Figure BDA0002540067800000091
After relaxation of r, solving the optimal flow controller mapping model to obtain a solution set
Figure BDA0002540067800000092
Figure BDA0002540067800000093
Wherein the content of the first and second substances,
Figure BDA0002540067800000094
is formed by flThe k mapping relationships of (a) and (b) form a set, the mapping relationships including mappings between switches and controllers and probabilities of the mappings; selecting the maximum value of k from the L k values as iteration times T; setting the path programmability of all offline streams to 0; will map the relationship
Figure BDA0002540067800000095
All are set to 0.
Wherein, in order to improve the detection efficiency, the solution set
Figure BDA0002540067800000096
The solutions in (1) may be sorted according to the size of the mapping probability value, for example, sorted from large to small according to the mapping probability value.
Step 3, establishing a set L to be tested containing all offline streams*(ii) a The value of the boolean variable IsMapped is set to 0.
The boolean variable IsMapped means whether the stream is recovered, and when the value is 1, the stream is already recovered, otherwise, the stream is not recovered.
In order to further increase the detection speed, the invention can establish a to-be-detected set L containing all offline streams during the next iteration according to the path programmability of the streams updated in the subsequent step 7*Then, the set L to be tested is put into*The flows in (1) are ordered in a way that the path programmability is from big to small.
Step 4, from the set L*Selecting an offline flow fl。
The selection mode of the offline stream may be random selection or may be according to the set L*The sorting order of (1).
Step 5, from flCorresponding to
Figure BDA0002540067800000097
To select the e-th mapping relation
Figure BDA0002540067800000098
Wherein e is more than or equal to 1 and less than or equal to k to obtain
Figure BDA0002540067800000099
The corresponding switch number i and the controller number j, i.e. the mapping relation is
Figure BDA00025400678000000910
Will be provided with
Figure BDA00025400678000000911
The value of (a) is set to 1; will be provided with
Figure BDA00025400678000000912
From
Figure BDA00025400678000000913
Deleting;
if it is
Figure BDA0002540067800000101
Satisfy the requirement of
Figure BDA0002540067800000102
And
Figure BDA0002540067800000103
then
Figure BDA0002540067800000104
To make the mapping feasible, flAccording to
Figure BDA0002540067800000105
Completing the mapping to switches and controllers and updating the total remaining processing power of all controllers in the network, calculating and updating flPath programmability of rlSetting the value of IsMapped as 1, and executing the step 6;
if it is
Figure BDA0002540067800000106
Not meet the requirements of
Figure BDA0002540067800000107
Or
Figure BDA0002540067800000108
Then will be
Figure BDA0002540067800000109
Setting to 0; if it is
Figure BDA00025400678000001010
If not, changing the value of e and executing the step 5; if it is
Figure BDA00025400678000001011
If it is empty, f islFrom L*And (4) deleting, and executing the step 7.
Wherein, from flCorresponding to
Figure BDA00025400678000001012
To select the e-th mapping relation
Figure BDA00025400678000001013
The method can be selected according to the principle that the mapping probability is prior, that is, the mapping relation with higher mapping probability is preferentially selected.
Step 6, if the total residual processing capacity is 0, the test is finished and the process is exited; otherwise, will flFrom L*Is deleted.
Step 7, if L*If not, executing step 4; otherwise, making T self-reduce by 1, if T is not 0, executing step 3 according to the path programmability of the updated flow; if T is 0, the test is finished and the process is exited.
The performance of the present embodiment was evaluated through experimental simulation, and the PG can increase the percentage of the recovered flow, while the programmability of the balance path is as high as 29%, and increase the total programmability of the recovered flow to 68%, which can reduce the communication overhead by 83% at most compared with the benchmark algorithm.
The above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims (3)

1. A method for optimizing the programmability of flow when multiple controllers fail in a software defined network is characterized by comprising the following steps:
step 1, establishing an optimal flow controller mapping model to describe a mapping relation between an offline flow and an online controller in a network, wherein the optimal flow controller mapping model is shown as the following formula:
Figure FDA0002969150470000011
Figure FDA0002969150470000012
Figure FDA0002969150470000013
Figure FDA0002969150470000014
Figure FDA0002969150470000015
Figure FDA0002969150470000016
where r is the minimum value of path programmability for all offline streams; l is the number of the offline stream f, and L is the total number of the offline streams in the network; i is the number of the offline switch, and N is the total number of the offline switches s in the network; j is the serial number of the online controller c, and M is the total number of the online controllers in the network;
Figure FDA0002969150470000017
Dijfor a switch siAnd a controller cjA propagation delay therebetween, λ is a constant greater than or equal to 0,
Figure FDA0002969150470000018
is flowed through siOff-line flow flThe number of paths contained on the offline switch s numbered i;
Figure FDA0002969150470000019
is a Boolean type variable when
Figure FDA00029691504700000110
When the value is 1, f is expressedlFlows through siAnd siAt least two paths to flOtherwise
Figure FDA00029691504700000111
The value is 0;
Figure FDA00029691504700000112
is a Boolean type variable when
Figure FDA00029691504700000119
A value of 1 indicates a flow through siF of (a)lMapping to cjWhen is coming into contact with
Figure FDA00029691504700000113
When the value is 0, the flow is indicated to pass through siF of (a)lNot mapped to cj；
Figure FDA00029691504700000114
Is cjThe remaining capacity of (c); max [. X [ ]]Is the maximum value;
step 2, mixing
Figure FDA00029691504700000115
After relaxation of r, solving the optimal flow controller mapping model to obtain a solution set
Figure FDA00029691504700000116
Figure FDA00029691504700000117
Wherein the content of the first and second substances,
Figure FDA00029691504700000118
is formed by flThe k mapping relationships of (a) and (b) form a set, the mapping relationships including mappings between switches and controllers and probabilities of the mappings; selecting the maximum value of k as iteration times T; setting the path programmability of all offline streams to 0; will map the relationship
Figure FDA0002969150470000021
All are set to 0;
step 3, establishing a set L to be tested containing all offline streams*；
Step 4, from the set L to be tested*Selecting an offline flow fl；
Step 5, from flCorresponding to
Figure FDA0002969150470000022
To select the e-th mapping relation
Figure FDA0002969150470000023
Wherein e is more than or equal to 1 and less than or equal to k to obtain
Figure FDA0002969150470000024
The corresponding switch number i and the controller number j, i.e. the mapping relation is
Figure FDA0002969150470000025
Will be provided with
Figure FDA0002969150470000026
The value of (a) is set to 1; will be provided with
Figure FDA0002969150470000027
From
Figure FDA0002969150470000028
Deleting;
if it is
Figure FDA0002969150470000029
Satisfy the requirement of
Figure FDA00029691504700000210
And
Figure FDA00029691504700000211
then
Figure FDA00029691504700000212
To make the mapping feasible, flAccording to
Figure FDA00029691504700000213
Completing the mapping to switches and controllers and updating the total remaining processing power of all controllers in the network, calculating and updating flPath programmability of rlExecuting the step 6;
if it is
Figure FDA00029691504700000214
Not meet the requirements of
Figure FDA00029691504700000215
Or
Figure FDA00029691504700000216
Then will be
Figure FDA00029691504700000217
Setting to 0; if it is
Figure FDA00029691504700000218
If not, changing the value of e and executing the step 5; if it is
Figure FDA00029691504700000219
Is emptyThen f will belFrom L*Deleting, and executing step 7;
step 6, if the total residual processing capacity is 0, the test is finished and the process is exited; otherwise, will flFrom L*Deleting, and executing step 7;
step 7, if L*If not, executing step 4; otherwise, making T self-reduce by 1, if T is not 0, executing step 3 according to the path programmability of the updated flow; if T is 0, the test is finished and the process is exited;
from f in said step 5lCorresponding to
Figure FDA00029691504700000220
To select the e-th mapping relation
Figure FDA00029691504700000221
And selecting according to a principle that the mapping probability is prior, namely preferentially selecting the mapping relation with higher mapping probability.
2. The method of claim 1, wherein in step 7, according to the updated path programmability of the stream, in the next iteration of step 3, the set L to be tested is first tested*The flows in the sequence are sequenced from large to small according to the programmability of the path; accordingly, the step 4 is from L*The offline stream with the greatest path programmability is selected.
3. The method of claim 1, wherein the path programmability is computed using a programmable path computation method based on a programmable fabric.
CN202010544094.4A 2020-06-15 2020-06-15 Method for optimizing programmability of flow when multiple controllers in software defined network fail Active CN111650878B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010544094.4A CN111650878B (en) 2020-06-15 2020-06-15 Method for optimizing programmability of flow when multiple controllers in software defined network fail

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010544094.4A CN111650878B (en) 2020-06-15 2020-06-15 Method for optimizing programmability of flow when multiple controllers in software defined network fail

Publications (2)

Publication Number Publication Date
CN111650878A CN111650878A (en) 2020-09-11
CN111650878B true CN111650878B (en) 2021-05-04

Family

ID=72341822

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010544094.4A Active CN111650878B (en) 2020-06-15 2020-06-15 Method for optimizing programmability of flow when multiple controllers in software defined network fail

Country Status (1)

Country Link
CN (1) CN111650878B (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1230090A (en) * 1997-10-01 1999-09-29 北方电讯有限公司 Communication system architecture and operating methods thereof
CN110391929A (en) * 2018-04-23 2019-10-29 深圳市格瑞信息科技有限公司 A kind of fault tolerant control method, device and fault-tolerant component
CN110908772A (en) * 2019-11-14 2020-03-24 北京理工大学 Energy-saving scheduling method for improving reliability of multiple workflows

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103675600B (en) * 2013-09-05 2016-03-30 国家电网公司 Based on the Fault Diagnosis of Distribution Network system and method for topological knowledge
CN110149226B (en) * 2019-05-14 2022-01-28 电子科技大学中山学院 Improved particle swarm algorithm for multi-controller deployment problem in software defined network

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1230090A (en) * 1997-10-01 1999-09-29 北方电讯有限公司 Communication system architecture and operating methods thereof
CN110391929A (en) * 2018-04-23 2019-10-29 深圳市格瑞信息科技有限公司 A kind of fault tolerant control method, device and fault-tolerant component
CN110908772A (en) * 2019-11-14 2020-03-24 北京理工大学 Energy-saving scheduling method for improving reliability of multiple workflows

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
SDN中控制器松弛管理及虚拟网络映射研究;李莹;《中国优秀硕士学位论文全文数据库信息科技辑》;20150630;全文 *
下一代网络技术中流量和资源管理机制研究;郭泽华;《中国博士学位论文全文数据库信息科技辑》;20150730;全文 *
面向聚类和回归的模糊求解技术及应用;刘解放;《中国博士学位论文全文数据库信息科技辑》;20180430;全文 *
高性能SDN数据面若干关键技术研究;郑凌;《中国博士学位论文全文数据库信息科技辑》;20200328;全文 *

Also Published As

Publication number Publication date
CN111650878A (en) 2020-09-11

Similar Documents

Publication Publication Date Title
CN109495300B (en) Reliable SDN virtual network mapping method
CN103703727B (en) The method and apparatus controlling the elastic route of business in split type architecture system
Vizarreta et al. Controller placement strategies for a resilient SDN control plane
CN105379196B (en) Method, system and computer storage medium for the routing of fault-tolerant and load balance
CN109768924B (en) SDN network multilink fault recovery method and system oriented to multi-stream coexistence
US8254263B2 (en) Method and apparatus for simplifying the computation of alternate network paths
US20140003228A1 (en) Optimizations in Multi-Destination Tree Calculations for Layer 2 Link State Protocols
CN107306224B (en) Routing path updating method, network management device and routing equipment
CN108011817A (en) A kind of method and system disposed again to power communication private network business route
CN104025513A (en) Hierarchy of control in a data center network
CN103716250A (en) IP Network resilient route optimization method based on load balancing
CN106533966A (en) Network service resource arranging method and apparatus
Guo et al. Improving the path programmability for software-defined WANs under multiple controller failures
Li et al. P4Resilience: Scalable resilience for multi-failure recovery in SDN with programmable data plane
CN113347102B (en) SDN link surviving method, storage medium and system based on Q-learning
Al-Rumaih et al. Spare capacity planning for survivable mesh networks
CN111650878B (en) Method for optimizing programmability of flow when multiple controllers in software defined network fail
US11706146B1 (en) Directing network traffic using local routing decisions with a global overview
CN109391488A (en) A kind of link management method and system for SDN network
Agarwal et al. Performance evaluation of hsrp, glbp and vrrp with interior gateway routing protocol and exterior gateway routing protocol
CN110213162A (en) Fault-tolerant routing method for large-scale computer system
CN107147539A (en) Judge that critical link provides the method and device that fast failure recovers in software defined network
Mota et al. Efficient routing table minimization for fault-tolerant irregular network-on-chip
Lemeshko et al. Application prospects of first hop redundancy protocols for fault-tolerant SDN controllers: a survey
CN109039681A (en) Method for optimizing route, storage device and the network equipment based on SDN

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant